A hierarchical Bayesian network approach for linkage disequilibrium modeling and data-dimensionality reduction prior to genome-wide association studies

BMC Bioinformatics

Research Article

Raphaël Mourad¹  Philippe Leray¹  Christine Sinoquet² 

[1] LINA, UMR CNRS 6241, Ecole Polytechnique de l'Université de Nantes, BP 50609, rue Christian Pauc, 44306, Nantes Cedex 3, France;LINA, UMR CNRS 6241, Université de Nantes, BP 92208, 2 rue de la Houssinie're, 44322, Nantes Cedex, France;

关键词: Linkage Disequilibrium;    Latent Variable;    Bayesian Network;    Directed Acyclic Graph;    Latent Class Model;   

DOI  :  10.1186/1471-2105-12-16

 received in 2010-04-15, accepted in 2011-01-12,  发布年份 2011

来源: Springer

【 摘 要 】

BackgroundDiscovering the genetic basis of common genetic diseases in the human genome represents a public health issue. However, the dimensionality of the genetic data (up to 1 million genetic markers) and its complexity make the statistical analysis a challenging task.ResultsWe present an accurate modeling of dependences between genetic markers, based on a forest of hierarchical latent class models which is a particular class of probabilistic graphical models. This model offers an adapted framework to deal with the fuzzy nature of linkage disequilibrium blocks. In addition, the data dimensionality can be reduced through the latent variables of the model which synthesize the information borne by genetic markers. In order to tackle the learning of both forest structure and probability distributions, a generic algorithm has been proposed. A first implementation of our algorithm has been shown to be tractable on benchmarks describing 105 variables for 2000 individuals.ConclusionsThe forest of hierarchical latent class models offers several advantages for genome-wide association studies: accurate modeling of linkage disequilibrium, flexible data dimensionality reduction and biological meaning borne by latent variables.

【 授权许可】

CC BY   
© Mourad et al; licensee BioMed Central Ltd. 2011